1. Field of the Invention
The present invention relates to the field of service applications and, more specifically, to monitoring the performance and activity of a server.
2. Description of the Prior Art
Monitoring the performance of a server enables administrators to determine if the server is working efficiently and to identify and correct problems affecting performance. For example, a database server may be monitored to identify worst-performing queries, to identify a cause of a deadlock, to monitor stored procedure performance, or to audit server activity.
Generally, a user creates a “template”, which is a collection of events and associated columns to be monitored. The user may then launch a trace of a server that requests information for the events and associated columns specified in the template. An event is an action generated within the server such as, for example, the start or end of a stored procedure, the start or end of a statement within a stored procedure, a lock acquired or released on a database object, or an error written to an error log. A column describes the data collected for an associated event. For example, a “binary data” column, when associated with a “lock acquired” event, may include an identifier of a locked page.
A problem related to server monitoring is that the traceable events and columns may vary depending on the type and version of the server to be monitored. Thus, conventional monitoring tools are designed to perform in combination with a particular server type and version. Specifically, conventional monitoring tools may be pre-programmed with all of the traceable events and columns for a corresponding server type and version. Multiple monitoring tools must, therefore, be acquired to perform in combination with different server types and versions.
Thus, there is a need in the art for a monitoring tool that is server and version independent. It is desired that such a tool have the ability to “discover” the traceable events and columns for a connected server and to request and receive a trace of such traceable events and columns.